1906-7. TRANSACTIONS. 135 



Speaking generally, mountains are not masses resting upon 

 the surface of the earth, but must be considered as masses im- 

 mersed in the earth, just as an iceberg is immersed in the water. 

 The greater the part that projects above the water, the greater 

 must be the part beneath the surface, for the amount of water 

 displaced must be equal to the floating mass, otherwise there 

 would not be equiUbrium. Somewhat similar it is with the moun- 

 tains. Were they resting on the surface, the stresses set up by 

 the superimposed mass would not only be enormous, but would 

 be greater than the crust could support. Furthermore as a super- 

 imposed mass it would materially affect the force of gravity in the 

 adjoining region. The most noted investigations of this question 

 was with reference to the attraction of the Himalayas in connec- 

 tion with the Great Trigonometrical Survey of India. Pendulum 

 observations have shown conclusively both in India and in Am- 

 erica that this is not the case. However complete equilibrium 

 or isostasy does not obtain, and hence the residual strains and 

 stresses. 



It is obvious how through meteoric agencies cycles of changes 

 are produced. The mountains by decrements are wandering sea- 

 ward, the continents are lightening, and the ocean bed is being 

 loaded, producing a deep-seated inflow from the sea towards the 

 land. These changes are continually taking place, the earth's 

 crust and surface are undergoing constant transformation, how- 

 ever minute; the strains and stresses are continually responding 

 to one another; vast rock formations that seem rigid are by the 

 slow process of time bent and contorted as if made of wire. But 

 when these responses are not synchronous, when there is a lag, 

 equilibrium can only be restored by rupture. This rupture will 

 be along the line of least resistance, and this is generally found in 

 a geological fault, an old rent in the crust, so well illustrated in the 

 California earthquake of last April. 



If the earth were a homogeneous body or if at least it were 

 composed of concentric shells each of homogeneous matter, then 

 the geodetic surveyor when carrying on large trigonometric surveys 

 would not be troubled with closing errors, other than those arising 

 from observations. There would be no error due to deflection 

 of the plumb line. As complete isostasy does not however exist, 

 these observed discordances, due to the unsymmetrical distribu- 

 tion of matter, are a measure of isostasy. 



Dr. J. F. Hayford has examined the data furnished by the 



